Facebook said a “faulty configuration change” was responsible for the widespread outages that lasted for more than six hours across its social media platforms on Monday.
In a statement on Tuesday Santosh Janardhan, vice president for engineering and infrastructure, said that configuration changes on the routers that co-ordinate communication between data centres caused this process to be interrupted.
He added: “This disruption to network traffic had a cascading effect on the way our data centres communicate, bringing our services to a halt.”
Facebook shares dropped 4.9% to $326.23 on the Nasdaq exchange in New York on Monday, although the US equity market suffered broad losses with the Nasdaq Composite Index falling 2.1% and the S&P 500 ending 1.3% lower. It was, however, Facebook’s biggest daily drop since November.
Janardhan’s statement added that the affected services – which included Facebook’s Messenger, WhatsApp and Instagram platforms – were back online and were in the process of being restored to regular operations.
He added: “We want to make clear at this time we believe the root cause of this outage was a faulty configuration change. We also have no evidence that user data was compromised as a result of this downtime.”
Earlier on Monday Mike Schroepfer, Facebook’s chief technology officer, apologised in a Twitter post saying the company was experiencing “networking issues”.
He added: “To every small and large business, family, and individual who depends on us, I’m sorry.”
Monday’s six-hour outage was not the social media network’s worst in terms of duration, but it had the biggest impact in terms of revenues lost and customers affected – given the company has grown, in size and reputation, into an industrial behemoth with more than 3.5 billion users of its services worldwide.
Facebook’s longest outage came in 2008, just four years after the company was founded, when a technical issue caused the platform to be offline for a whole day.
The last major outage was on 13 March, 2019, lasting several hours – again, following a “configuration change” on the company’s servers.